Thursday, May 14, 2009

Racism, Eggs and Visualization

One of my favorite books is The Logic of Life: A Rational Economics in an Irrational World by Tim Harford. Not only do I like the style of economics in the book, but I admire the way Tim Harford writes. I think that Harford communicates the findings of economics beautifully. For anyone who appreciates good logic and clear exposition, I highly recommend anything produced by Tim Harford: speeches, books, blog writings, etc. Harford just exudes an ideal outlook on life and that shines through in his writing.

He has now started producing video shorts. These video shorts are a wonderful way to get the ideas of economics out into public discourse. They're all posted on his webpage at

My favorite video short describes Thomas Schelling's model of racial segregation. Of course, I think the model is interesting, but what I find most interesting is the difference in exposition. To see what I mean, let's give it a try. Imagine you have a chessboard in front of you. Read the following description of how extreme racial segregation can occur even if each member of society has only a mild preference for his own race:
Lay out alternating black and white pieces, remove any twenty, and then add five just to mix things up a bit. The board now represents a mixed neighborhood.

Now, these black pieces aren't extreme racists. They're happy to live in a mixed neighborhood, but they don't want their white neighbors to outnumber their black neighbors more than 2 to 1. The white pieces feel exactly the same way. So, take any piece that is outnumbered by more than 2 to 1 and move it to the closet acceptable location.

When you do this, you'll find something astonishing. The black pieces and white pieces will separate out like oil and vinegar. Even a mild preference for the color of your neighbor can lead to extreme segregation.

When you're done, watch the video (here). Now, that's a cool video, don't you think?

It is striking how well the video conveys Schelling's chessboard experiment. Some ideas just have to be seen to be understood.

1 comment:

  1. Jim RC gave a talk a couple of years ago about some research he did in this area... He developed a cool animation that showed these types of trends in Urban areas over time (with the ability to adjust the neighbor type 'preference' variable). He also did a similar one with female advancement in corporate America. R can do cool stuff :)


Please feel free to share your ideas about this post in the open forum. Be mindful that comments in this blog are moderated. Please keep your comments respectful and on point.